import pandas as pd
import numpy as np
import seaborn as sns
import plotly.express as px
from matplotlib import pyplot as plt
%matplotlib inline


#html export
import plotly.io as pio
pio.renderers.default = 'notebook'

df = pd.read_excel(r'Case Study Data - Read Only.xlsx')
df

# make column names and values uniform
df.columns = df.columns.str.lower().str.replace(' ', '_')

categorical_columns = df.dtypes[df.dtypes == 'object'].index
for c in categorical_columns:
    df[c] = df[c].str.lower().str.replace(' ', '_')

df

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 333405 entries, 0 to 333404
Data columns (total 7 columns):
 #   Column               Non-Null Count   Dtype         
---  ------               --------------   -----         
 0   date                 333405 non-null  datetime64[ns]
 1   anonymized_category  333405 non-null  object        
 2   anonymized_product   333405 non-null  object        
 3   anonymized_business  333405 non-null  object        
 4   anonymized_location  333405 non-null  object        
 5   quantity             333405 non-null  int64         
 6   unit_price           333397 non-null  float64       
dtypes: datetime64[ns](1), float64(1), int64(1), object(4)
memory usage: 17.8+ MB

# duplicates
df.duplicated().sum()

3524

#dropping duplicates
df = df.drop_duplicates()

df.isna().sum()

date                   0
anonymized_category    0
anonymized_product     0
anonymized_business    0
anonymized_location    0
quantity               0
unit_price             8
dtype: int64

df[df.unit_price.isna()]

df_product_3d7fdf = df[df['anonymized_product'] == 'product-3d7f']
df_product_3d7fdf

df_product_3d7fdf['unit_price'].unique()

array([860., 875., 885., 870.,  nan])

df_product_3d7fdf[df_product_3d7fdf['unit_price'] == 860.0]

df_product_3d7fdf[df_product_3d7fdf['anonymized_category'] == 'category-94']

# dropping nulls
df = df.dropna()
df.info()

<class 'pandas.core.frame.DataFrame'>
Index: 329873 entries, 0 to 333404
Data columns (total 7 columns):
 #   Column               Non-Null Count   Dtype         
---  ------               --------------   -----         
 0   date                 329873 non-null  datetime64[ns]
 1   anonymized_category  329873 non-null  object        
 2   anonymized_product   329873 non-null  object        
 3   anonymized_business  329873 non-null  object        
 4   anonymized_location  329873 non-null  object        
 5   quantity             329873 non-null  int64         
 6   unit_price           329873 non-null  float64       
dtypes: datetime64[ns](1), float64(1), int64(1), object(4)
memory usage: 20.1+ MB

df

# since column 'date' is already a datetime object,

# ------------Create the 'Month-Year' column---------------#
df2 = df.copy() # creating a copy for original cleaned df

# Create a 'Month-Year' column, keeping the date as the first of the month
df2['month-year'] = df['date'].dt.to_period('M').dt.to_timestamp()

# Format it for display, but still keep the datetime
df2['month-year'] = df2['month-year'].dt.strftime('%B %Y')

# Drop the 'date' column
df2.drop(columns=['date'], inplace=True)

# Move 'Month-Year' column to the first position
columns = ['month-year'] + [col for col in df2.columns if col != 'month-year']
df2 = df2[columns]

# Display the updated DataFrame to check the new column
df2

# Calculate the 'Value' column (quantity × unit_price)
df3 = df2.copy()
df3['Value'] = df3['quantity'] * df3['unit_price']
df3

# Group by anonymized_category
category_sales = df3.groupby('anonymized_category').agg(
    Total_Quantity=('quantity', 'sum'),
    Total_Value=('Value', 'sum')
).reset_index()

category_sales

# graphing total sales and total value per anonymized category
import plotly.graph_objects as go

# bar chart for Total Quantity
bar = go.Bar(
    x=category_sales["anonymized_category"],
    y=category_sales["Total_Quantity"],
    name="Total Quantity",
    marker=dict(color="steelblue")
)

# line chart for Total Value
line = go.Scatter(
    x=category_sales["anonymized_category"],
    y=category_sales["Total_Value"],
    name="Total Value",
    mode="lines+markers",  # Line with markers
    line=dict(color="firebrick", width=3),  # Line color and width
    marker=dict(size=8),  # Marker size
    yaxis="y2"  # Assign to secondary y-axis
)

# Combine both traces
fig = go.Figure(data=[bar, line])

# Update layout for dual y-axes
fig.update_layout(
    title="Sales Overview by Category",
    xaxis_title="Category",
    yaxis=dict(
        title="Total Quantity",
        titlefont=dict(color="steelblue"),  # Match bar chart color
        tickfont=dict(color="steelblue")  # Match bar chart color
    ),
    yaxis2=dict(
        title="Total Value",
        titlefont=dict(color="firebrick"),  # Match line chart color
        tickfont=dict(color="firebrick"),  # Match line chart color
        overlaying="y",  # Overlay on the same plot
        side="right"  # Position on the right
    ),
    legend_title="Metrics",
    barmode="group",  # Group bars together
    template="plotly_white"  # Cleaner theme
)

# Show the figure
fig.show()

# Group by anonymized_business
business_sales = df3.groupby('anonymized_business').agg(
    Total_Quantity=('quantity', 'sum'),
    Total_Value=('Value', 'sum')
).reset_index()

business_sales

# Line chart for Total Value (green)
bar_value = go.Line(
    x=business_sales["anonymized_business"],
    y=business_sales["Total_Value"],
    name="Total Value",
    marker=dict(color="green"),
)

# layout
layout_value = go.Layout(
    title="Total Value by Business",
    xaxis_title="Business",
    yaxis_title="Total Value",
    template="plotly_white",
)

# figure
fig_value = go.Figure(data=[bar_value], layout=layout_value)

# Show the plot
fig_value.show()

c:\Program Files\Python312\Lib\site-packages\plotly\graph_objs\_deprecations.py:378: DeprecationWarning:

plotly.graph_objs.Line is deprecated.
Please replace it with one of the following more specific types
  - plotly.graph_objs.scatter.Line
  - plotly.graph_objs.layout.shape.Line
  - etc.

# line chart for Total Quantity (blue)
bar_quantity = go.Line(
    x=business_sales["anonymized_business"],
    y=business_sales["Total_Quantity"],
    name="Total Quantity",
    marker=dict(color="blue"),
)

# layout
layout_quantity = go.Layout(
    title="Total Quantity by Business",
    xaxis_title="Business",
    yaxis_title="Total Quantity",
    template="plotly_white",
)

# figure
fig_quantity = go.Figure(data=[bar_quantity], layout=layout_quantity)

# Show the plot
fig_quantity.show()

# Create a copy of df
df4 = df.copy()

# Add 'Value' column
df4['value'] = df4['quantity'] * df4['unit_price']

# Create a new column 'Month-Year' in df4 (not df)
df4['month-year'] = df4['date'].dt.to_period('M').dt.to_timestamp()

# Format 'Month-Year' to display as "August 2024"
df4['month-year-formatted'] = df4['month-year'].dt.strftime('%B %Y')

# Reorder columns to make 'Month-Year-Formatted' the first column
df4 = df4[['month-year-formatted'] + [col for col in df4.columns if col != 'month-year-formatted']]

del df4['date']

df4.info()

<class 'pandas.core.frame.DataFrame'>
Index: 329873 entries, 0 to 333404
Data columns (total 9 columns):
 #   Column                Non-Null Count   Dtype         
---  ------                --------------   -----         
 0   month-year-formatted  329873 non-null  object        
 1   anonymized_category   329873 non-null  object        
 2   anonymized_product    329873 non-null  object        
 3   anonymized_business   329873 non-null  object        
 4   anonymized_location   329873 non-null  object        
 5   quantity              329873 non-null  int64         
 6   unit_price            329873 non-null  float64       
 7   value                 329873 non-null  float64       
 8   month-year            329873 non-null  datetime64[ns]
dtypes: datetime64[ns](1), float64(2), int64(1), object(5)
memory usage: 25.2+ MB

# Group the data by 'Month-Year' and calculate the monthly average of 'Quantity' and 'Value'
monthly_avg_sales = df4.groupby('month-year').agg({'quantity': 'mean', 'value': 'mean'}).reset_index()

# Convert 'Month-Year' to a string for plotting purposes
monthly_avg_sales['month-year'] = monthly_avg_sales['month-year'].astype(str)

# Group the data by 'Month-Year' and calculate the monthly averages
monthly_avg_sales = df4.groupby('month-year').agg({'quantity': 'mean', 'value': 'mean'}).reset_index()

# Convert 'Month-Year' to a string for plotting
monthly_avg_sales['month-year'] = monthly_avg_sales['month-year'].astype(str)

# Create a line plot with dual y-axes
fig = go.Figure()

# Add the line for Value
fig.add_trace(go.Scatter(
    x=monthly_avg_sales['month-year'], 
    y=monthly_avg_sales['value'], 
    name='Average Value', 
    line=dict(color='red'),
    yaxis='y1'
))

# Add the line for Quantity
fig.add_trace(go.Scatter(
    x=monthly_avg_sales['month-year'], 
    y=monthly_avg_sales['quantity'], 
    name='Average Quantity', 
    line=dict(color='green'),
    yaxis='y2'
))

# Update layout for dual y-axes
fig.update_layout(
    title="Monthly Averages of Sales (Quantity and Value)",
    xaxis=dict(title="Month-Year"),
    yaxis=dict(
        title="Average Value",
        titlefont=dict(color="blue"),
        tickfont=dict(color="blue"),
    ),
    yaxis2=dict(
        title="Average Quantity",
        titlefont=dict(color="green"),
        tickfont=dict(color="green"),
        anchor="x",
        overlaying="y",
        side="right",
    ),
    legend=dict(x=0.5, y=1.1, orientation="h"),
)

# Show the plot
fig.show()

# Group the data by 'Month-Year' and calculate the monthly sum of 'Quantity' and 'Value'
monthly_sum_sales = df4.groupby('month-year').agg({'quantity': 'sum', 'value': 'sum'}).reset_index()

# Convert 'Month-Year' to a string for plotting purposes
monthly_sum_sales['month-year'] = monthly_sum_sales['month-year'].astype(str)

# Group the data by 'Month-Year' and calculate the monthly averages
monthly_sum_sales = df4.groupby('month-year').agg({'quantity': 'sum', 'value': 'sum'}).reset_index()

# Convert 'Month-Year' to a string for plotting
monthly_sum_sales['month-year'] = monthly_sum_sales['month-year'].astype(str)

# Create a line plot with dual y-axes
fig = go.Figure()

# Add the line for Value
fig.add_trace(go.Scatter(
    x=monthly_sum_sales['month-year'], 
    y=monthly_sum_sales['value'], 
    name='Total Sum Value', 
    line=dict(color='red'),
    yaxis='y1'
))

# Add the line for Quantity
fig.add_trace(go.Scatter(
    x=monthly_sum_sales['month-year'], 
    y=monthly_sum_sales['quantity'], 
    name='Total Sum Quantity', 
    line=dict(color='green'),
    yaxis='y2'
))

# Update layout for dual y-axes
fig.update_layout(
    title="Monthly Sum of Sales (Quantity and Value)",
    xaxis=dict(title="Month-Year"),
    yaxis=dict(
        title="Total Value",
        titlefont=dict(color="blue"),
        tickfont=dict(color="blue"),
    ),
    yaxis2=dict(
        title="Total Quantity",
        titlefont=dict(color="green"),
        tickfont=dict(color="green"),
        anchor="x",
        overlaying="y",
        side="right",
    ),
    legend=dict(x=0.5, y=1.1, orientation="h"),
)

# Show the plot
fig.show()

# top 5 most frequently purchased products (based on Quantity).

# Group by product and sum the quantities
top_products = df4.groupby('anonymized_product')['quantity'].sum().reset_index()

# Sort in descending order and select the top 5
top_5_products = top_products.sort_values(by='quantity', ascending=False).head(5)

# Display the result
top_5_products

# Create a bar graph
fig = px.bar(
    top_5_products,
    x='anonymized_product',
    y='quantity',
    title='Top 5 Most Purchased Products (Quantity)',
    labels={'anonymized_product': 'Product', 'quantity': 'Total Quantity'},
    text='quantity'  # Display the quantity on the bars
)

# Customize the layout
fig.update_layout(
    xaxis_title='Product',
    yaxis_title='Total Quantity',
    template='plotly_white'
)

# Show the plot
fig.show()

# top 5 most valuable products (based on Value).

# Group by product and sum the quantities
top_value_products = df4.groupby('anonymized_product')['value'].sum().reset_index()

# Sort in descending order and select the top 5
top_value_products = top_value_products.sort_values(by='value', ascending=False).head(5)

# Display the result
top_value_products

# Create a bar graph
fig = px.bar(
    top_value_products,
    x='anonymized_product',
    y='value',
    title='Top 5 Most Valuable Products (Value)',
    labels={'anonymized_product': 'Product', 'value': 'Total Quantity'},
    text='value'  # Display the quantity on the bars
)

# Customize the layout
fig.update_layout(
    xaxis_title='Product',
    yaxis_title='Total Value',
    template='plotly_white'
)

# Show the plot
fig.show()

df4

# Group by Anonymized Business and calculate required metrics
business_segmentation = df4.groupby('anonymized_business').agg(
    total_quantity=('quantity', 'sum'),
    total_value=('value', 'sum'),
    transaction_frequency=('anonymized_business', 'count')
).reset_index()

# Sorting for analysis (optional)
top_businesses = business_segmentation.sort_values(by='total_value', ascending=False)
top_businesses

# scatter plot to analyze relationships between total quantity, total value, and transaction frequency
fig = px.scatter(
    business_segmentation,
    x='total_quantity',
    y='total_value',
    size='transaction_frequency',
    color='transaction_frequency',
    title='Customer Segmentation Based on Purchasing Behavior',
    labels={
        'total_quantity': 'Total Quantity Purchased',
        'total_value': 'Total Value Contributed ($)',
        'transaction_frequency': 'Transaction Frequency'
    },
    hover_data=['anonymized_business']
)

# Show the plot
fig.show()

business_segmentation

# trying to see the distribution of total_value to calssify the businesses
fig = px.box(business_segmentation, y='total_value')

# Show the figure
fig.show()

# Calculate quartiles
Q1 = business_segmentation['total_value'].quantile(0.25)
Q2 = business_segmentation['total_value'].quantile(0.50)
Q3 = business_segmentation['total_value'].quantile(0.75)

# Compute IQR and upper whisker (excluding extreme outliers)
IQR = Q3 - Q1
upper_whisker = Q3 + 1.5 * IQR

# Define function to categorize values into quartiles
def categorize(value):
    if value <= Q1:
        return "0 - Q1"
    elif Q1 < value <= Q2:
        return "Q1 - Q2"
    elif Q2 < value <= Q3:
        return "Q2 - Q3"
    elif Q3 < value <= upper_whisker:
        return "Q3 - Q4"
    else:
        return "Outlier"

# Apply function to categorize values
business_segmentation['quartile_category'] = business_segmentation['total_value'].apply(categorize)

# Ensure quartile categories are ordered
quartile_order = ["0 - Q1", "Q1 - Q2", "Q2 - Q3", "Q3 - Q4", "Outlier"]

# Convert to categorical with ordered=True
business_segmentation['quartile_category'] = pd.Categorical(
    business_segmentation['quartile_category'],
    categories=quartile_order,
    ordered=True
)

# Count occurrences, now sorted correctly
quartile_counts = business_segmentation['quartile_category'].value_counts().sort_index()

# Display results
print("Count of values in each quartile:\n")
print(quartile_counts)

Count of values in each quartile:

quartile_category
0 - Q1     1200
Q1 - Q2    1200
Q2 - Q3    1200
Q3 - Q4     526
Outlier     674
Name: count, dtype: int64

# Get the minimum and maximum values
min_value = business_segmentation['total_value'].min()
max_value = business_segmentation['total_value'].max()

# Display results
print(f"Minimum Value: {min_value}")
print(f"First Quartile (Q1): {Q1}")
print(f"Second Quartile (Median/Q2): {Q2}")
print(f"Third Quartile (Q3): {Q3}")
print(f"Fourth Quartile (Upper Whisker): {upper_whisker}")
print(f"Maximum Value: {max_value}")

Minimum Value: 0.0
First Quartile (Q1): 9207.5
Second Quartile (Median/Q2): 41869.0
Third Quartile (Q3): 234042.0
Fourth Quartile (Upper Whisker): 571293.75
Maximum Value: 28037358.0

# Define function to classify businesses based on their total_value
def classify_business(value):
    if value > Q3:
        return "high"
    elif Q2 < value <= Q3:
        return "medium"
    else:
        return "low"

# Apply the classification function to the dataframe
business_segmentation['value_category'] = business_segmentation['total_value'].apply(classify_business)

business_segmentation

# forecasting with arima
import plotly.graph_objects as go
from statsmodels.tsa.arima.model import ARIMA

# Ensure the Month-Year column is in datetime format
df4['month-year'] = pd.to_datetime(df4['month-year'])

# Aggregate total sales by month
monthly_sales = df4.groupby('month-year')['value'].sum()

# Fit an ARIMA model (order can be optimized based on ACF/PACF)
model = ARIMA(monthly_sales, order=(2,1,2))  
model_fit = model.fit()

# Forecast for the next 3 months
forecast = model_fit.forecast(steps=3)

# Generate future date range
future_dates = pd.date_range(start=monthly_sales.index[-1], periods=4, freq='M')[1:]

# Convert dates to "Month-Year" format (e.g., "January 2024")
formatted_dates = future_dates.strftime('%B %Y')

# Create Plotly figure
fig = go.Figure()

# Add historical sales data
fig.add_trace(go.Scatter(
    x=monthly_sales.index.strftime('%B %Y'),  # Format historical x-axis labels
    y=monthly_sales.values,
    mode='lines',
    name='Historical Sales',
    line=dict(color='blue')
))

# Add forecasted sales data
fig.add_trace(go.Scatter(
    x=formatted_dates, 
    y=forecast.values,
    mode='lines+markers',
    name='Forecasted Sales',
    line=dict(color='red', dash='dash')
))

# Update layout
fig.update_layout(
    title="3-Month Sales Forecast",
    xaxis_title="Month-Year",
    yaxis_title="Total Sales (Value)",
    xaxis=dict(tickmode='array', tickvals=monthly_sales.index.strftime('%B %Y')),
    template="plotly_white"
)

# Show the plot
fig.show()

c:\Program Files\Python312\Lib\site-packages\statsmodels\tsa\base\tsa_model.py:473: ValueWarning:

No frequency information was provided, so inferred frequency MS will be used.

c:\Program Files\Python312\Lib\site-packages\statsmodels\tsa\base\tsa_model.py:473: ValueWarning:

No frequency information was provided, so inferred frequency MS will be used.

c:\Program Files\Python312\Lib\site-packages\statsmodels\tsa\base\tsa_model.py:473: ValueWarning:

No frequency information was provided, so inferred frequency MS will be used.

c:\Program Files\Python312\Lib\site-packages\statsmodels\tsa\statespace\sarimax.py:966: UserWarning:

Non-stationary starting autoregressive parameters found. Using zeros as starting parameters.

c:\Program Files\Python312\Lib\site-packages\statsmodels\tsa\statespace\sarimax.py:978: UserWarning:

Non-invertible starting MA parameters found. Using zeros as starting parameters.

C:\Users\Deninjo\AppData\Local\Temp\ipykernel_10284\468917369.py:18: FutureWarning:

'M' is deprecated and will be removed in a future version, please use 'ME' instead.

# Identify any unusual spikes or drops in sales performance (Quantity or Value)
monthly_sum_sales = df4.groupby('month-year').agg({'quantity': 'sum', 'value': 'sum'}).reset_index()

# Convert 'Month-Year' to a string for plotting purposes
monthly_sum_sales['month-year'] = monthly_sum_sales['month-year'].astype(str)

# Group the data by 'Month-Year' and calculate the monthly averages
monthly_sum_sales = df4.groupby('month-year').agg({'quantity': 'sum', 'value': 'sum'}).reset_index()

# Convert 'Month-Year' to a string for plotting
monthly_sum_sales['month-year'] = monthly_sum_sales['month-year'].astype(str)

# Create a line plot with dual y-axes
fig = go.Figure()

# Add the line for Value
fig.add_trace(go.Scatter(
    x=monthly_sum_sales['month-year'], 
    y=monthly_sum_sales['value'], 
    name='Total Sum Value', 
    line=dict(color='red'),
    yaxis='y1'
))

# Add the line for Quantity
fig.add_trace(go.Scatter(
    x=monthly_sum_sales['month-year'], 
    y=monthly_sum_sales['quantity'], 
    name='Total Sum Quantity', 
    line=dict(color='green'),
    yaxis='y2'
))

# Update layout for dual y-axes
fig.update_layout(
    title="Monthly Sum of Sales (Quantity and Value)",
    xaxis=dict(title="Month-Year"),
    yaxis=dict(
        title="Total Value",
        titlefont=dict(color="blue"),
        tickfont=dict(color="blue"),
    ),
    yaxis2=dict(
        title="Total Quantity",
        titlefont=dict(color="green"),
        tickfont=dict(color="green"),
        anchor="x",
        overlaying="y",
        side="right",
    ),
    legend=dict(x=0.5, y=1.1, orientation="h"),
)

# Show the plot
fig.show()

df4

# correlation matrix between nmerical columns
X = df4[[ 'quantity', 'unit_price','value']]

# Calculate the correlation matrix
correlation_matrix = X.corr()

# Create a heatmap using seaborn
plt.figure(figsize=(10, 8))
sns.heatmap(correlation_matrix, annot=True, fmt=".2f", cmap="coolwarm", cbar=True)

# Add a title
plt.title("Correlation Heatmap of Selected Features", fontsize=16)

# Show the heatmap
plt.show()

# Group by Anonymized Category and calculate required metrics
category_segmentation = df4.groupby('anonymized_category').agg(
    total_quantity=('quantity', 'sum'),
    total_value=('value', 'sum'),
    transaction_frequency=('anonymized_category', 'count')
).reset_index()

# Sorting for analysis (optional)
top_categories = category_segmentation.sort_values(by='total_value', ascending=False)
top_categories.head()

# scatter plot to analyze relationships between total quantity, total value, and transaction frequency
fig = px.scatter(
    category_segmentation,
    x='total_quantity',
    y='total_value',
    size='transaction_frequency',
    color='transaction_frequency',
    title='Category Segmentation Based on Purchasing Behavior',
    labels={
        'total_quantity': 'Total Quantity Purchased',
        'total_value': 'Total Value Contributed ($)',
        'transaction_frequency': 'Transaction Frequency'
    },
    hover_data=['anonymized_category']
)

# Show the plot
fig.show()

#  Identify businesses that have reduced their purchase frequency over time

# Convert to datetime if not already
df4['month-year-formatted'] = pd.to_datetime(df4['month-year-formatted'])

# Format month-year as "Month Year" (e.g., "August 2024")
df4['month-year-formatted'] = df4['month-year-formatted'].dt.strftime('%B %Y')

# Group by business and formatted month-year, count purchases in each month
purchase_trend = df4.groupby(['anonymized_business', 'month-year-formatted']).size().reset_index(name='purchase_count')

# Calculate the trend in purchase frequency per business
purchase_trend['prev_purchase_count'] = purchase_trend.groupby('anonymized_business')['purchase_count'].shift(1)
purchase_trend['change'] = purchase_trend['purchase_count'] - purchase_trend['prev_purchase_count']

# Identify businesses with a declining trend
declining_customers = purchase_trend[purchase_trend['change'] < 0]

declining_customers

C:\Users\Deninjo\AppData\Local\Temp\ipykernel_10284\855654940.py:2: UserWarning:

Could not infer format, so each element will be parsed individually, falling back to `dateutil`. To ensure parsing is consistent and as-expected, please specify a format.

# Summary Statistics on Purchase Change
declining_customers['change'].describe()

count    7267.000000
mean      -10.530893
std        13.789538
min      -310.000000
25%       -13.000000
50%        -6.000000
75%        -3.000000
max        -1.000000
Name: change, dtype: float64

# Businesses with the Biggest Declines
top_decliners = declining_customers.nsmallest(25, 'change')
top_decliners

# Trend Over Time (Monthly Summary)

# Ensure 'month-year-formatted' is in datetime format
declining_customers['month-year-formatted'] = pd.to_datetime(declining_customers['month-year-formatted'], format='%B %Y')

# Group by month-year and sum changes
monthly_decline = declining_customers.groupby('month-year-formatted')['change'].sum()

# Sort by date (ensures chronological order)
monthly_decline = monthly_decline.sort_index(ascending=True)

# Convert back to string format for proper labeling
monthly_decline.index = monthly_decline.index.strftime('%B %Y')

# Create an interactive bar chart using Plotly
fig = px.bar(
    x=monthly_decline.index,
    y=monthly_decline.values,
    labels={'x': 'Month-Year', 'y': 'Total Decline in Purchases'},
    title="Total Monthly Decline in Purchases",

)

# Improve layout
fig.update_layout(
    xaxis_title="Month-Year",
    yaxis_title="Total Decline",
    xaxis=dict(tickangle=-45),  # Rotate x-axis labels for better readability
    bargap=0.3  # Adjust bar spacing
)

# Show the interactive graph
fig.show()

C:\Users\Deninjo\AppData\Local\Temp\ipykernel_10284\4292986156.py:4: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

	DATE	ANONYMIZED CATEGORY	ANONYMIZED PRODUCT	ANONYMIZED BUSINESS	ANONYMIZED LOCATION	QUANTITY	UNIT PRICE
0	2024-08-18 21:32:00	Category-106	Product-21f4	Business-de42	Location-1ba8	1	850.0
1	2024-08-18 21:32:00	Category-120	Product-4156	Business-de42	Location-1ba8	2	1910.0
2	2024-08-18 21:32:00	Category-121	Product-49bd	Business-de42	Location-1ba8	1	3670.0
3	2024-08-18 21:32:00	Category-76	Product-61dd	Business-de42	Location-1ba8	1	2605.0
4	2024-08-18 21:32:00	Category-119	Product-66e0	Business-de42	Location-1ba8	5	1480.0
...	...	...	...	...	...	...	...
333400	2024-01-09 20:49:00	Category-119	Product-e98d	Business-f9ff	Location-1979	1	1770.0
333401	2024-08-20 20:24:00	Category-106	Product-21f4	Business-72bc	Location-689f	1	850.0
333402	2024-08-20 20:24:00	Category-76	Product-6e9c	Business-72bc	Location-689f	1	2575.0
333403	2024-10-10 18:19:00	Category-91	Product-523e	Business-2693	Location-b27b	1	1520.0
333404	2024-10-10 18:19:00	Category-75	Product-b31e	Business-2693	Location-b27b	1	4420.0

	date	anonymized_category	anonymized_product	anonymized_business	anonymized_location	quantity	unit_price
0	2024-08-18 21:32:00	category-106	product-21f4	business-de42	location-1ba8	1	850.0
1	2024-08-18 21:32:00	category-120	product-4156	business-de42	location-1ba8	2	1910.0
2	2024-08-18 21:32:00	category-121	product-49bd	business-de42	location-1ba8	1	3670.0
3	2024-08-18 21:32:00	category-76	product-61dd	business-de42	location-1ba8	1	2605.0
4	2024-08-18 21:32:00	category-119	product-66e0	business-de42	location-1ba8	5	1480.0
...	...	...	...	...	...	...	...
333400	2024-01-09 20:49:00	category-119	product-e98d	business-f9ff	location-1979	1	1770.0
333401	2024-08-20 20:24:00	category-106	product-21f4	business-72bc	location-689f	1	850.0
333402	2024-08-20 20:24:00	category-76	product-6e9c	business-72bc	location-689f	1	2575.0
333403	2024-10-10 18:19:00	category-91	product-523e	business-2693	location-b27b	1	1520.0
333404	2024-10-10 18:19:00	category-75	product-b31e	business-2693	location-b27b	1	4420.0

	date	anonymized_category	anonymized_product	anonymized_business	anonymized_location	quantity	unit_price
108112	2024-07-03 17:53:00	category-94	product-3d7f	business-4fce	location-f37d	2	NaN
150961	2024-12-16 18:33:00	category-79	product-dfc8	business-8bbf	location-3fc0	1	NaN
151142	2024-12-22 14:42:00	category-122	product-15e0	business-c575	location-1979	3	NaN
272379	2024-06-27 12:15:00	category-92	product-ccbc	business-14b6	location-1979	1	NaN
278284	2024-08-14 21:09:00	category-101	product-84a5	business-4be1	location-bb69	21	NaN
278384	2024-12-30 14:17:00	category-95	product-15f3	business-1a74	location-f37d	1	NaN
310385	2024-03-31 14:03:00	category-114	product-9204	business-c9dc	location-689f	1	NaN
327152	2024-08-13 16:20:00	category-107	product-7eed	business-0d61	location-1ba8	1	NaN

	date	anonymized_category	anonymized_product	anonymized_business	anonymized_location	quantity	unit_price
196	2024-09-25 13:05:00	category-94	product-3d7f	business-cdc1	location-e2f8	1	860.0
287	2024-11-13 19:36:00	category-94	product-3d7f	business-7f77	location-8959	1	860.0
319	2024-11-01 16:54:00	category-94	product-3d7f	business-2b91	location-3fc0	1	860.0
372	2024-07-11 11:20:00	category-94	product-3d7f	business-2b24	location-66f4	5	875.0
447	2024-12-09 19:35:00	category-94	product-3d7f	business-1b52	location-7f37	1	885.0
...	...	...	...	...	...	...	...
332771	2024-10-15 20:01:00	category-94	product-3d7f	business-0ea3	location-4ea1	1	860.0
332802	2024-10-21 19:04:00	category-94	product-3d7f	business-5415	location-128a	1	860.0
332842	2024-09-20 13:43:00	category-94	product-3d7f	business-5760	location-689f	1	860.0
333002	2024-10-08 17:15:00	category-94	product-3d7f	business-8603	location-b27b	1	860.0
333188	2024-11-08 21:20:00	category-94	product-3d7f	business-8bbf	location-3fc0	1	860.0

	date	anonymized_category	anonymized_product	anonymized_business	anonymized_location	quantity	unit_price
196	2024-09-25 13:05:00	category-94	product-3d7f	business-cdc1	location-e2f8	1	860.0
287	2024-11-13 19:36:00	category-94	product-3d7f	business-7f77	location-8959	1	860.0
319	2024-11-01 16:54:00	category-94	product-3d7f	business-2b91	location-3fc0	1	860.0
565	2024-09-15 15:52:00	category-94	product-3d7f	business-f749	location-3fc0	2	860.0
780	2024-10-03 15:08:00	category-94	product-3d7f	business-cdc1	location-e2f8	1	860.0
...	...	...	...	...	...	...	...
332771	2024-10-15 20:01:00	category-94	product-3d7f	business-0ea3	location-4ea1	1	860.0
332802	2024-10-21 19:04:00	category-94	product-3d7f	business-5415	location-128a	1	860.0
332842	2024-09-20 13:43:00	category-94	product-3d7f	business-5760	location-689f	1	860.0
333002	2024-10-08 17:15:00	category-94	product-3d7f	business-8603	location-b27b	1	860.0
333188	2024-11-08 21:20:00	category-94	product-3d7f	business-8bbf	location-3fc0	1	860.0

Kwanza Tukule Data Analysis¶

Importing Libraries¶

Loading Dataset¶

Attributes¶

Data Cleaning and Preparation¶

Summary of Issues Identified and Steps Taken to Resolve Them:¶

Feature Engineering:
Creating the column: “Month-Year” (e.g., August 2024) from the “DATE” column.

EDA (Exploratory Data Analysis)¶

Sales Overview¶

Trends Over Time¶

Performance analysis¶

Advanced Analysis¶

Perform a segmentation analysis of businesses (Anonymized Business) based on their purchasing behavior:¶

Classifying businesses into 3 groups (e.g., High Value, Medium Value, Low Value)¶

Business Value Segmentation & Engagement Strategies¶

Forecasting¶

Anomaly Detection¶

Correlation Analysis¶

Strategic Insights and Recommendations¶

Product Strategy¶

Customer Retention¶

Open-Ended Problem¶

	month-year	anonymized_category	anonymized_product	anonymized_business	anonymized_location	quantity	unit_price
0	August 2024	category-106	product-21f4	business-de42	location-1ba8	1	850.0
1	August 2024	category-120	product-4156	business-de42	location-1ba8	2	1910.0
2	August 2024	category-121	product-49bd	business-de42	location-1ba8	1	3670.0
3	August 2024	category-76	product-61dd	business-de42	location-1ba8	1	2605.0
4	August 2024	category-119	product-66e0	business-de42	location-1ba8	5	1480.0
...	...	...	...	...	...	...	...
333398	November 2024	category-121	product-898d	business-0e99	location-689f	5	1360.0
333401	August 2024	category-106	product-21f4	business-72bc	location-689f	1	850.0
333402	August 2024	category-76	product-6e9c	business-72bc	location-689f	1	2575.0
333403	October 2024	category-91	product-523e	business-2693	location-b27b	1	1520.0
333404	October 2024	category-75	product-b31e	business-2693	location-b27b	1	4420.0

	anonymized_category	Total_Quantity	Total_Value
0	category-100	76824	134902751.0
1	category-101	19564	35577822.0
2	category-102	1786	464463.0
3	category-104	1217	1557598.0
4	category-105	1579	2690719.0
5	category-106	6521	5932763.0
6	category-107	2729	4170797.0
7	category-108	9756	5101375.0
8	category-109	1446	1263226.0
9	category-110	10529	5483386.0
10	category-111	6715	4387343.0
11	category-113	741	1254083.0
12	category-114	3	8600.0
13	category-115	348	425360.0
14	category-116	856	422745.0
15	category-117	5	1550.0
16	category-118	21	7560.0
17	category-119	68332	103454819.0
18	category-120	169715	319178743.0
19	category-121	14669	22327643.0
20	category-122	1223	3493480.0
21	category-123	286	730730.0
22	category-124	4	10060.0
23	category-125	123	297060.0
24	category-74	941	1927871.0
25	category-75	151330	544658700.0
26	category-76	71719	344939553.0
27	category-77	28455	76741382.0
28	category-78	9766	9792609.0
29	category-79	2215	1184953.0
30	category-81	142	72061.0
31	category-82	4759	3930818.0
32	category-83	2436	4039483.0
33	category-84	11933	6798158.0
34	category-85	22997	33762533.0
35	category-86	8	3320.0
36	category-89	238	136850.0
37	category-90	15	15750.0
38	category-91	20853	44152103.0
39	category-92	6953	10468723.0
40	category-94	23668	16750815.0
41	category-95	4116	7466932.0
42	category-96	1427	2249424.0
43	category-97	2711	2628309.0
44	category-98	2152	2519695.0
45	category-99	1964	1589480.0

	anonymized_business	Total_Quantity	Total_Value
0	business-0000	8	10445.0
1	business-0005	1	2645.0
2	business-0029	26	77340.0
3	business-003d	98	221761.0
4	business-0072	127	225056.0
...	...	...	...
4795	business-ffa9	3	6740.0
4796	business-ffae	6	10530.0
4797	business-ffb1	266	438115.0
4798	business-ffd2	37	67723.0
4799	business-ffff	110	110285.0

	anonymized_product	quantity
338	product-66e0	46957
750	product-e805	42602
476	product-8f75	37566
127	product-29ee	35940
213	product-4156	28487

	anonymized_product	value
750	product-e805	262787281.0
476	product-8f75	158797460.0
338	product-66e0	70704225.0
127	product-29ee	68248274.0
213	product-4156	56956007.0

	anonymized_business	total_quantity	total_value	transaction_frequency
2906	business-978e	13991	28037358.0	2614
4781	business-fe7d	6743	26997121.0	653
1873	business-6068	8214	16464195.0	1474
144	business-07de	6065	16258068.0	1160
2353	business-7a03	6318	13968451.0	1147
...	...	...	...	...
2616	business-8654	1	55.0	1
3666	business-c19b	1	55.0	1
1478	business-4b87	5	10.0	1
4060	business-d78d	1	2.0	1
4619	business-f712	22	0.0	1

	anonymized_business	total_quantity	total_value	transaction_frequency	quartile_category	value_category
0	business-0000	8	10445.0	8	Q1 - Q2	low
1	business-0005	1	2645.0	1	0 - Q1	low
2	business-0029	26	77340.0	6	Q2 - Q3	medium
3	business-003d	98	221761.0	31	Q2 - Q3	medium
4	business-0072	127	225056.0	101	Q2 - Q3	medium
...	...	...	...	...	...	...
4795	business-ffa9	3	6740.0	3	0 - Q1	low
4796	business-ffae	6	10530.0	5	Q1 - Q2	low
4797	business-ffb1	266	438115.0	105	Q3 - Q4	high
4798	business-ffd2	37	67723.0	22	Q2 - Q3	medium
4799	business-ffff	110	110285.0	107	Q2 - Q3	medium

	anonymized_business	month-year-formatted	purchase_count	prev_purchase_count	change
3	business-0000	November 2024	1	2.0	-1.0
11	business-0072	December 2024	10	23.0	-13.0
13	business-0072	June 2024	8	14.0	-6.0
14	business-0072	May 2024	3	8.0	-5.0
17	business-0072	September 2024	10	22.0	-12.0
...	...	...	...	...	...
20675	business-ffd2	June 2024	3	8.0	-5.0
20676	business-ffd2	March 2024	2	3.0	-1.0
20678	business-ffd2	September 2024	4	5.0	-1.0
20681	business-ffff	October 2024	25	48.0	-23.0
20682	business-ffff	September 2024	18	25.0	-7.0

	anonymized_business	month-year-formatted	purchase_count	prev_purchase_count	change
12637	business-978e	February 2024	115	425.0	-310.0
17010	business-cf0d	November 2024	26	313.0	-287.0
16745	business-cb1f	September 2024	219	451.0	-232.0
2941	business-245e	July 2024	25	196.0	-171.0
10187	business-78a8	November 2024	6	160.0	-154.0
9081	business-6baf	January 2024	2	155.0	-153.0
8047	business-6011	September 2024	18	164.0	-146.0
12645	business-978e	September 2024	278	415.0	-137.0
3006	business-24c3	September 2024	37	171.0	-134.0
4763	business-3955	July 2024	27	149.0	-122.0
13826	business-a8bd	March 2024	23	141.0	-118.0
525	business-07de	December 2024	6	114.0	-108.0
534	business-07de	September 2024	111	216.0	-105.0
7188	business-5677	September 2024	106	210.0	-104.0
3147	business-263c	December 2024	1	104.0	-103.0
10341	business-7ab4	November 2024	24	125.0	-101.0
15483	business-bc52	September 2024	34	129.0	-95.0
5891	business-468e	November 2024	59	151.0	-92.0
10271	business-7a03	February 2024	18	110.0	-92.0
18098	business-de42	March 2024	17	107.0	-90.0
13421	business-a3fe	September 2024	47	136.0	-89.0
18907	business-e997	December 2024	43	130.0	-87.0
2525	business-1f55	November 2024	17	103.0	-86.0
8076	business-6068	March 2024	47	133.0	-86.0
2713	business-2197	February 2024	7	91.0	-84.0

Kwanza Tukule Data Analysis¶

Importing Libraries¶

Loading Dataset¶

Attributes¶

Data Cleaning and Preparation¶

Summary of Issues Identified and Steps Taken to Resolve Them:¶

Feature Engineering: Creating the column: “Month-Year” (e.g., August 2024) from the “DATE” column.

EDA (Exploratory Data Analysis)¶

Sales Overview¶

Trends Over Time¶

Performance analysis¶

Advanced Analysis¶

Perform a segmentation analysis of businesses (Anonymized Business) based on their purchasing behavior:¶

Classifying businesses into 3 groups (e.g., High Value, Medium Value, Low Value)¶

Business Value Segmentation & Engagement Strategies¶

Forecasting¶

Anomaly Detection¶

Correlation Analysis¶

Strategic Insights and Recommendations¶

Product Strategy¶

Customer Retention¶

Open-Ended Problem¶

Feature Engineering:
Creating the column: “Month-Year” (e.g., August 2024) from the “DATE” column.